The Agency Cost of Concentrated Institutional Ownership: Evidence from Indonesia

The purpose is to analyze the role of institutional ownership in firms with managerial ownership in reducing agency costs for the Indonesian listed firm. The polled data from the annual report of companies listed on the In-donesia Stock Exchange and data obtained are pooled from 949 units from 2008 to 2020. We have analyzed through regression of data from dependent is agency cost, independent managerial ownership, annual sales, total assets, independent commissionaires, and control variable is debt ratio. The total sales to total assets ratio has been used as a proxy for agency costs. The result is that higher agency cost proves more effectively (more than 1) in us-ing assets to generate annual sales. Quartile analysis has been used to determine the interval of managerial ownership, the result is that managerial ownership at the lower level (34%-59%) and higher (60%-74%) is significant to agency costs. Inverse, when none of the managerial ownership, much lower (<33%), and much higher (>74%) is insignificant to agency cost. Our results support the predictions of Jensen & Meckling (1976), which state that ownership structure as voting power and monitoring are mechanisms for aligning the interests of shareholders. The presence of much lower of managerial ownership results in managers not having enough power to produce agency costs. No different, when managerial ow-nership is at a much higher level, they also do not have the potential to produce agency costs because institutional investors can monitor more op-portunistic management behavior. Both levels of managerial ownership pro-ve more effectively use assets to generate annual sales.Originality/value is a study focusing on one of the largest emerging economies, i.e. Indonesia.


INTRODUCTION
In the seminal paper of Jensen & Meckling (1976), separation of ownership and control produces agency costs when a manager owns less than 100 percent equity. Whenever a manager controls an outside investor's investment other than their own, then agency costs are a natural consequence of being paid by shareholders due to utility maximization by self-interested managers (Lin & Chang, 2008;Rashid, 2016). The sacrifice of wealth by shareholders and potential costs because managers do not bear the full consequences of actions are agency costs. When the manager sells a fraction of the equity, has effective control, and has the incentive and ability to do on-the-job perks, shirk, and maximize self-interest, the shareholder's wealth is reduced.
Managerial ownership is an internal mechanism to prevent and limit managers (McConaughy, 1995;Shleifer & Vishny, 1989; RM Stulz, 1999). The manager aligns with the interest of the shareholders when managers feel that they are a fraction of the firm, and agency costs are reduced (Villalonga et al., 2015). This is because the manager has incentives to reduce underinvestment and enforce company policy restrictions (Jensen & Meckling, 1976;Muñoz Mendoza et al., 2021;Shleifer & Vishny, 1989). Furthermore, an increase in more concentrated ownership stimulates managers to have incentives to force company management to avoid monitoring. This action increases agency costs, called the entrenchment hypothesis (Berger et al., 2005;DeAngelo et al., 2006). In contrast, higher managerial ownership effectively reduces agency costs because it is more operationally efficient (RenéM M. Stulz, 1990). The difference in managers' ownership of the equity fraction produces variable and unsolved agency costs. Scholten (2005) especially ensures an external control mechanism other than managerial ownership, namely the market for corporate control that can prevent and limit actions that are not aligned with shareholders when ownership is low. High non-manager shareholder ownership incentivizes them to take action to maximize value, and managers do not have to lose their jobs. This way, variations in the market for corporate control reduce moral hazard and adverse selection problems by managers (Coffee, 2001). More convincingly, non-managers are institutional investors with greater monitoring of resources and voting power (Chaudhary, 2022). Thus, agency costs can vary depending on the nature of ownership, namely who owns the firm, i.e., managers/directors or institutional investors as external shareholders. In a sole proprietorship, manager-ownership has zero agency costs. No firm listed on the national stock exchange is 100 percent equity-owned by a manager, and there is no zero agency cost (Jensen & Meckling, 1976). Thus, the ownership structure is measured by the manager's fraction of equity and the institutional ownership's presence.
Our model predicts the degree of relationship presence of institutional ownership in a firm owned by the manager in agency costs. Concentrated institutional ownership provides more monitoring capabilities and voting power in company decisions, impacting agency costs. The theory has suggested, and empirical evidence confirms, that institutional investors can provide active monitoring for more passive or less-informed investors (Almazan et al., 2005). The presence of fund managers among institutional investors results in more effective monitoring than individual investors (Bebchuk et al., 2017). Pressure-resistant investors have big claims at stake; thus, monitoring is carried out more professionally and actively to produce their investment targets (Chaudhary, 2022). Therefore, large institutional investors have more bargaining in each company to have a large, non-negligible effect on shareholders' voting. Institutional investors such as banks and financial institutions with more significant resources produce better capabilities than individuals. In turn, it limits the manager's discretion, so it is possible to prevent them from sometimes making inefficient decisions in the firm (Chaudhary, 2022).
In contrast, when institutional ownership is diluted, the profits from the actions of each shareholder will be shared among other shareholders, and it is unrealistic to concentrate ownership when they invest and are rationally apathetic (Bebchuk et al., 2017). We use the ratio of annual sales to total assets (hereafter, ASTA) as a proxy for agency costs (Ang et al., 2000). ASTA has an inverse relationship with agency cost and manager effectiveness of the firm in using its assets when the agency cost increases. Our paper structure consists of a literature review and hypothesis construction in the second part. The third section describes the research method. Section four presents the results obtained from the data analysis. The conclusion is given in the last section. The relationship between ownership structure and governance challenges depends on the concept of dispersion with concentrated ownership (Fama & Jensen, 1983;La Porta et al., 1999). More governance is needed when the ownership structure is dispersed. Therefore, better management is required for shareholders with lower ownership and the limited capacity and ability to monitor managers unless they have incentives and a market for corporate control. In contrast, when ownership is concentrated, the agency shifts into a majority and minority shareholder conflict.
A survey by De La Cruz et al. (2019) in 54 capital markets at the end of 2017 categorizes investors as private corporations and holding companies, public sector, strategic individuals and family members, institutional investors, and other free-float, including retail investors. All types of investors may have different incentives in varying public equity holdings. In turn, it influences them to allocate their capital (market capitalization) to listed firms and how they monitor performance and participate in firms' decision-making.
In companies listed on the Indonesian Stock Exchange, 36 percent of all firms have equity allocations from private corporations, 20 percent from the public sector, 12 percent from strategic individuals, 11 percent from institutional ownership, and 21 percent from retail investors. More than 70 percent of companies are owned by single shareholders (private corporations) who hold more than half of the equity capital. Concentrated ownership worsens when using the definition of institutional investors from The Indonesia Central Securities Depository (KSEI), where corporations are added to institutional investors (Prosperindo, 2020). The ownership structure model may vary, such as the dispersion found in the United States, so managers' actions are monitored through information disclosure, legal protection, corporate control or takeover market, and managers' lack of equity. However, in several continental European countries, companies are monitored by majority shareholders, such as banks (Chaudhary, 2022). Consequently, agency costs in continental European countries are more absolute than in the United States. In Indonesia, an emerging economy, it is possible to find a mixture of agency costs between traditional conflicts (principal-agent) with concentrated ownership. Regulations in Indonesia affirm the presence of the board of commissioners in the ownership structure, which aims to oversee managers' actions and prevent expropriation by the majority to minority shareholders (IFC, 2014). Due to highly concentrated ownership, lack of takeover regulations, inefficient markets, and significant transaction costs associated with the takeover process, a market for corporate control mechanism is needed.

Managerial Ownership: Internal Mechanism to Reduce Agency Cost
We discuss the substitution between managerial and institutional ownership variations in agency conflict. Managers, like everyone else, have many different personal goals and ambitions that they pursue. One of those goals may be the desire to act in the interests of shareholders, but other than that, managers may also want wealth, empire-building, or something completely different (Shleifer & Vishny, 1989). Inversely, shareholders as investors do not manage the company directly and only care about getting rich from their equity (Cariola et al., 2011).
Additional literature on agents (managers) more informed than principals (external shareholders) has been described by Akerlof (1970) in the used-car market. The presence of asymmetric information makes sellers more informed than buyers. As a result, "lemons" (low-quality cars) are offered more than "plums" (high-quality cars). It is no different when the equity ownership by shareholders is less than 100 percent and handing over the firm's management to the manager. As a result, firms like this are subject to agency costs due to the shareholders' lack of time or ability to observe managers fully. As suggested, separating ownership and control functions from shareholders (Fama & Jensen, 1983) may not be effective. So agency costs are still expected in companies with single ownership when the manager is an outsider.
Firms may have two extremes of separation of ownership and control functions. First, the owner has 100 percent equity ownership, can internalize all monitoring costs, and has the right to hire or fire managers. Agency costs may be low because shareholders receive the same benefits as monitoring costs are 100 percent. Conversely, the manager has no equity and is paid as an employee. Both are impossible in a public firm (Ang et al., 2000), So our focus is that managers own some fraction but not all of the firm's equity. Agency costs arise when engaging managers are not aligned with shareholders and take the form of a preference for on-the-job perks, shirking, and entrenchment, which reduces shareholder wealth. Managers with superior information can take action in the form of empire-building (Jensen & Meckling, 1976), entrenchment (Shleifer & Vishny, 1989), and over-confidence (Stein, 2001).
A model by Jensen & Meckling (1976), which suggests a trade-off in agency cost between firms with less or more insider ownership, has motivated us to investigate further. Agency costs are produced whenever managers control investments from outside investors in addition to their own because there is a fundamental conflict. When the manager's equity ownership is close to 0 percent, the manager will likely have incentives to deviate from the shareholders' (non-manager) goals (Levitas et al., 2011;Lin & Chang, 2008). As a consequence, agency costs increase. In contrast, an increase in more concentrated ownership (close to 100 percent) stimulates managers to have incentives to force company management to avoid monitoring from the capital market. As a result, agency costs increase, which is called the entrenchment hypothesis (Berger et al., 2005;DeAngelo et al., 2006). In contrast, when a manager's share ownership is lower, they are incentivized to align themself with the majority shareholders (Ang et al., 2000).
Capital market control is believed to be a mechanism to reduce agency conflict by the manager (Shleifer & Vishny, 1989). The takeover threat helps curb conflict because managers know they will be replaced if they perform poorly.
The hypothesis: firms with less or more managerial ownership have higher agency costs than moderate managerial ownership.

Institutional Ownership: External Mechanism to Reduce Agency Cost
The adverse selection problem arises when a manager withholds valuable information about their abilities to incentivize the manager to increase their value to the firm. It also makes manager dismissal very costly for shareholders. The manager can make investment decisions to control businesses, although it may not be the most efficient decision for the firms (Shleifer & Vishny, 1989). Even though the market for corporate control, such as a takeover, results in managers being forced to take actions following the interests of shareholders (Scholten, 2005). Mark (2016) summarizes several opinions that if the market for corporate control is allowed to function when managers are afraid of a takeover, it produces job losses. Although shareholders do not have direct control, the incompetence of managers will be removed through a takeover process.
In organizations where decision agents do not bear most of the wealth effect of their decisions, the board of directors is used as a controlling device to ensure that managers act efficiently and in the interests of their shareholders. The board of directors always has the power to hire, fire, and compensate top-level decision managers and to ratify and monitor important decisions (Claessens et al., 2000;Fama & Jensen, 1983). In Indonesia, councils can exercise these rights by law, but this is rarely done (Setyahadi & Narsa, 2020). Even when the board of directors does not elect managers, it usually does not have enough information to judge whether the project is maximizing value or not. Acquiring this knowledge can be very costly for the board of directors.
Theoretically, monitoring is defined as observing agent effort or the results achieved through supervision, consequently producing agency costs that shareholders must pay. Monitoring costs by shareholders is suggested to prevent and limit agent behavior from detrimental actions (Jensen & Meckling, 1976;Temel-Candemir, 2005). The presence of the majority shareholder has more significant benefits than the minority. They obtain benefits greater than the costs of obtaining information when asymmetric information is present. In addition, when voting, they get a large enough share of the vote (even with minor ownership) so that it can effectively affect the company's results when it is found that managers are not aligned with the interests of shareholders. Minority shareholders experience this difficulty, although they can collectively internalize it because of the possibility of different interests.
Institutional investors such as banks and financial institutions with greater resources produce better capabilities than individuals. In turn, it limits the manager's discretion, so it is possible to prevent them from sometimes making inefficient decisions in the firm (Jensen & Meckling, 1976). Firms with institutional ownership have fewer agency costs than those without institutional ownership (John & Litov, 2010).
The Hypothesis: concentrated ownership of institutional investors has greater ability and resources to monitor, resulting in lower ASTA effectiveness than dispersed institutional ownership.

RESEARCH METHOD
The data source is the annual report of companies listed on the Indonesia Stock Exchange, and data obtained are pooled from 949 units of observation from 2008 to 2020. Companies with equity-owned solely by managers account for 9.06%, whereas the other 90.94% are owned by managers together with institutional ownership. We eliminate the finance sector from the sample because of the different characteristics of the agency conflict (Maher & Anderson, 1999) and the differences in regulators in Indonesia (Martono et al., 2020), with the results in Table 1.

Variable
Dependent Variable: This paper uses the annual sales to total assets as a proxy for agency costs ASTA (Ang et al., 2000;Chaudhary, 2022). This ratio is inversely proportional to agency costs. Firms with a higher ratio (more than 1) have more effective use of company assets in generating annual sales than a lower ratio of firms (less than 1). The reason is that higher sales cannot be claimed as a measure of welfare for shareholders because the cash flow generated from higher sales can be used by management to increase their allowances rather than increase the company's value (Chaudhary, 2022).
Explanatory variables: Institutional ownership is the percentage of shares owned by seven types of institutional ownership, namely corporations, mutual funds, securities companies, insurance, pension funds, financial institutions, and foundations, according to the Indonesian Central Securities Depository (Prosperindo, 2020). Managerial ownership is the percentage of directors' shares (Rashid, 2016). Jensen & Mecking's model finds a trade-off from agency costs between more or less managerial ownership. We loosen the assumption of bonding and residual costs because agency conflicts can naturally be eliminated as long as the principals can fully observe agents' behavior. As a fraction of the owned firm, the institution has the resources to monitor the agent's behavior more.
Control Variable: We use the proposition from Jensen & Meckling (1976) that firms with outside equity and the presence of debt financing affect managers' incentives for moral hazard problems through risk-shifting. The resulting costs of preventive action can be considered agency costs. Inversely, debt can discipline managers' resource use behavior (Jensen, 1986). Institutional ownership has a bigger role in monitoring when debt produces more agency costs because of the moral hazard and discipline mechanism. The role of debt in determining agency costs increases as Indonesia is a bank-based country (Warjiyo, 2015). Indonesia has a two-tier system, placing independent commissioners more in supervising managers, where majority shareholders are affiliated with directors to take action to damage minority shareholders (Claessens et al., 2000). We eliminate the role of the independent commissioner in the model to invest more in the conflict between managers with institutional ownership, not the majority with the minority.

Data analysis
We divide agency costs into two parts: firms owned by the manager and firms owned by managerinstitutional investors. The ownership function to ratify and monitor is more optimal when implemented through voting power (Fama & Jensen, 1983). Monitoring results can be executed and influence in-firm decisions when institutional investors have majority ownership. We divide the level of institutional ownership into quantiles, namely much lower, lower, higher, and much higher (Brandon, 2018).
Data analysis includes a univariate test to determine whether there are differences in agency costs between firms owned by the manager and owned by manager institutions. Thus, we first compare the mean agency costs of the two investigated groups using an independent t-test. To confirm that our finding is robust concerning sample distributions, we also perform nonparametric tests on the difference between the medians through the Wilcoxon test (Ang et al, 2000). Second, we predict agency costs with regression analysis at each level of institutional ownership. Table 1, panel A presents descriptive statistics for the hypothesized variables explaining agency costs. Columns 2 and rows 1, 3, and 4 show the number of observations and the average and median ratio of annual sales to total assets. Consistent with our previous expectation, namely concentrated institutional ownership, the number of firms owned by managers and institutional investors (863 or 90.94% of observation units) is more than the number of firms without institutional ownership (82 or 9.06% of 949 observation units). Thus, it generates enough observations in these two groups to compare these ASTA ratios statistically. We compare firms owned by the manager and institutional ownership in the much lower, lower, higher, and much higher categories, which correspond to equity ownership of less than 33 percent, 34-59 percent, 60 -74 percent, and more than 74 percent. Panel B, columns 4 and 7 find that firms with much lower and much higher institutional ownership are significant and robust. The presence of much institutional ownership as shareholders in firms with owned managers (34 percent) produces more effectiveness in using its assets of 37 percent than without institutional ownership. As a result, the firm's effectiveness in using its assets is increased by 2,144 (in billion Rupiah) in capturing annual sales (37 percent of 5,722). When institutional ownership is much higher, it increases the average total assets of 1,261 in obtaining yearly sales (28 percent of 4,505) from higher institutional ownership. Of course, the present value of the agency's cost of equity is undoubtedly several times higher.

Univariate Test
We verify directly that agency costs are generated due to less or more insider ownership (Jensen & Meckling, 1976). Institutional ownership is inversely proportional to managerial ownership -much lower institutional ownership has quite a high managerial ownership and continues to decrease when institutional ownership increases. Managerial ownership is 34% and 3%, and the mean agency cost is 1.04 and 1.25 in firms with much lower institutional ownership and much higher institutional ownership. Concen-trated institutional ownership produces more effectiveness in using its assets than concentrated managerial ownership. Both types of institutional ownership have the firm's effectiveness in using its assets more than 1. Managers have more incentives to produce annual sales that exceed total assets. We exclude the extreme case where manager ownership is 100 percent equity and zero percent equity. Thus, the level of managerial ownership varies from 0 to 100 percent, producing differences in agency costs. Therefore, managers align with shareholders' interest on less or more insider ownership.
Interestingly, we suspect that when managerial ownership is low enough, the average is 34% (not enough voting power), which produces a market for corporate control in increasing effectiveness in using its assets. This is partly because achieving the voice needed to make corporate management decisions is difficult. No different, when institutional ownership is concentrated (presence by the fund manager), effectiveness in using its assets is achieved through monitoring. The regression test to clarify the univariate is presented in table 2, Panel B.  Table 2, Panel B examines firms with no institutional ownership (column 2). When managers have relatively low company equity, an average of 31 percent, managers do not have enough voting power in company decisions. The classic problem in the separation of ownership and control is that the personal vote of the manager will be counted little or not at all in the meeting. Manager vote will practically be an alternative or not considered because there is too little ownership (Coffee, 2001). Indonesian Corporate Law (ICL) distinguishes rights between individual and collective shareholders related to the decision-making process and firm organization (IFC, 2014). More specifically, the rights of shareholders based on the percentage of common stock ownership are attending and voting in the General Meeting of Shareholders (GMS); receiving payment of dividends and remainder of assets from liquidation; exercising other right under the ICL. Various regulations in Indonesia allow controlling shareholders from a group of minority shareholders to have more voting power than individual shareholders. As a result, the voting power of equity owned can control the company directly and indirectly to (1) determine the appointment or dismissal of directors or commissioners; or (2) amend the articles of association of the public company, the company (Kairupan, 2009 We do not divide institutional investors like the model by Almazan et al. (2005), where they are divided into active institutional (more skilled employees, the ability to collect more information, faced with fewer regulatory and legal issues) and passive institutional investors. The Indonesia Central Securities Depository (KSEI) classification of institutional investors is used, i.e., corporations, mutual funds, securities companies, insurance, pension funds, financial institutions, and foundations (Prosperindo, 2020).
Columns 3 to 6 explain when much lower institutional ownership in a firm owned by a manager produces an ASTA ratio of less than one or less effectiveness of the firm in using its assets at 34% equity held by the manager (negative and significant). Increasing institutional ownership to moderate produces more effectiveness for the firm in using its assets by the manager. In contrast, when much higher institutional ownership results in managers with shares, it does not determine the effectiveness of ASTA. It produces interesting conflicts whenever managers are managed by outside investors other than their own. This situation results in managers only bearing the cost of a smaller fraction, thus only bearing a small part of their actions. The manager's actions can provide financial benefits, including monetary payments, higher salaries, and empire-building. The manager also gets non-pecuniary benefits, including non-monetary facilities. It is worse when institutional ownership is much lower and cannot prevent managers from voting. They seem to let managers act on their behalf.
Thus, when managers have more equity than institutional ownership, asymmetric information produces low ASTA effectiveness. Dispersion of ownership exacerbates the situation, where each shareholder has different interests. The manager bears only a fraction of the cost of a fraction of the equity, and there may be no difference in benefits when the ownership is 100 percent (Jensen & Meckling, 1976).
Columns 4 and 5 explain institutional ownership rising together with the disappearance of management ownership in the moderate category (lower and higher), producing more effectiveness for the firm in using its assets. Institutional ownership has a more significant monitoring role than the individual. Institutional investors have big claims at stake, so they monitor more professionally and actively to achieve their profits (Reed, 2003). These institutions have investment interests, and, more importantly, each institutional investor has a fund manager who manages their funds; in other words, the monitoring carried out by these institutional investors benefits themselves (Bebchuk et al., 2017).

Regression Analysis
Institutional ownership from lower to higher produces more effectiveness of ASTA even though firms with lower institutional have "capital" total assets greater than higher institutional ownership. We suspect that the increase in equity ownership by institutions allows an increase in voting power to implement monitoring results. Institutional intervention is significant when the total assets are greater, and more agency conflict is produced through the increase in total assets (Ang et al., 2000). More specifically, Bebchuk et al. (2017) developed ownership theory based on private benefits extracted by controlling shareholders.
We believe that when the potential for personal gain is more (more total assets), the institution is incentivized to take control to earn profits and block the possibility of outsiders' takeover. A survey by Shleifer & Vishny (1989) adds to our insight into why some countries have a low concentration of ownership (e.g., US) and some high (e.g., Europe).

Results Discussion
This section estimates how variation in ownership structure is essential for agency costs. Agency theory has predicted that when management owns less than 100 percent of equity, shareholders pay the agency costs resulting from shirking and consumption of perquisites by managers (Jensen & Meckling, 1976). In the dispersed type of ownership, without a controlling (majority) shareholder, commissioners and directors are incentivized to dictate their successors. Since dispersed shareholders can thus be expected to be rationally apathetic, managers will be relatively constrained in their actions through the market for corporate control. This way, the market for corporate control substitutes for monitoring and reduces managers' asymmetric information and the moral hazard problem (Easterbrook, 1984). The basic idea is that the stock market offers a mechanism to take over underperforming firms and thus provide a way to replace existing management with more efficient ones or merge firms with more efficient rivals, even in a dispersed ownership structure. Efficiency here, of course, is understood from the perspective of shareholders. When the mechanism for disciplining managers together with performance is not suitable, the company's takeover in a competitive market (the market for corporate control) becomes one of the most effective ways for shareholders to eliminate managers who do not maximize value (Shleifer & Vishny, 1989). The role of the market for corporate control is substituted for monitoring by other shareholders (Scholten, 2005).
When institutional ownership is more concentrated without managerial ownership, it produces more effectiveness for the firm in using its assets. Our findings strengthen the theory from Bebchuk et al. (2017) that institutions with better monitoring capabilities and more equity have a completely different incentive. Institutional investors generally enter into contracts with investment advisors (fund managers) to manage portfolios of investment funds. Investment managers acting on behalf of institutional investors need support for the safety of investment by firms. The increase in voting through ownership can be used in GMS regarding the selection of directors and intervention in the company's internal processes. Investment managers must secure their investment assets by voting and voting based on the clarity of their information. Institutional investors participate in corporate voting (Aghion et al., 2013).

CONCLUSION
In this article, we use data from listed firms in IDX for 2008 -2020 to examine the difference in agency costs between firms owned by a manager with and without institutional ownership. Companies in Indonesia are more than 90 percent owned by institutional investors and are inversely proportional to managerial ownership. Manager as minority ownership and institutional presence as majority shareholder produce more effectiveness of the firm in using its assets.
We also examine the determinants of agency costs in the regression framework. Our results support the predictions of Jensen & Meckling (1976), which state that ownership structure as voting power and monitoring are mechanisms for aligning the interests of shareholders. First, managers cannot make decisions because of the low power of equity, so the allocation of total assets in capturing annual sales is determined by institutional ownership. Second, it produces more effectiveness for the firm in using its assets because the fund manager of institutional investors (majority ownership) acts for them, not the interests of the firm whose equity is owned (Bebchuk et al., 2017). Thus, the counterproductive mechanism produces more effectiveness for the firm in using its assets when institutional owners, through fund managers, intervene more with managers who own some shares. Our results have an important implication that the sample companies have business relationships with institutional ownership, and we then leave the substance of these relationships for future research agendas.